SYMPTOMATIC

BY B. N. TEBBS, M.A., M.D., B.C.CANTAB.

Received June 6th-Read October 25th, 1904.

SY3MPTOMATIC, seconldary, or mnetastatic parotitis is a subject round which mnuch interest and discussion have centred. A complication of other diseases, rather than a disease szti generis, it is best known, perhaps, as a sequela of abdominal operations and disorders, and of certain of the specific fevers, but there are many other conditions of which it is a recognised, if less common, complicationl. The only systematic study of the subject in this country is contained in the admirable papers by Mr. Stephen Paget (1) published in the years 1886 and 1887, which conistitute the chief source of our knowledge of this dis- order. Series of cases have also been published in America by Goodell and Morley (2), and in France by Benoit. The present paper is founded on a collection of 77 hitherto unpublished cases of this disorder, nearly all of which are taken from the records of St. George's Hospital. I must express my indebtedness to the members of the staff of that hospital for their kind permission to use these cases, anid to the other gentlemen who have kindly communicated cases to me. I have referred to other published cases but have not included them in the series. Anl analysis of the 77 cases will be found at the end of this paper. Previous observers have, I think, been rather inclined to VOL. LXXXVIII. 4 36 SYMPTOMATIC PAROTITIS look on the disorder as mnore or less of a specific disease, and to endeavour to adopt one explanation that will cover all the cases. Believing symptomatic parotitis to be a disease of rather heterogeneous pathogeny, I have attempted to classify the cases provisionally as follows: A. Acute parotitis following or in connection with (1) abdominal operations and diseases of the alimrentary cainal and its appendages; (2) operation on and disturbances of the generative organs; (3) operation on or injury of portions of the body other than the abdominal and pelvic viscera; (4) certain diseases of metabolism and chronic intoxications; (5) the specific fevers; (6) inflammatory conditions of neighbouring parts. B. Recurrent parotitis and chronic enlargement of the gland. The acute cases seem to fall naturally into two main groups, one of which is undoubtedly of septic and the other probably of toxic origin. Group 4, and probably some cases in group 3, represenit the toxic parotitis, and the other groups the septic variety. Some of the recurrent cases appear to depend largely, if not wholly, on vaso-motor disturba,nces, and would constitute a separate group. Believing that in the majority of the septic cases the infection of the gland arises not by a spread along the duct from the mouth but by the blood stream, I have paid special attention in the cases to the evidence of a primary septic focus, and to the co-existence of other septic com- plications in the patients that develop parotitis. I propose to treat these groups one by one, and finally to take up the consideration of certain general points in connection with the disorder. In the case of post-operative parotitis I have endeavoured to arrive at some idea of its relative frequency after various operations, by taking the total number of such operations performed durinig a period of fourteen years, and working out the percentage of cases followed by this complication. Cases fatal within forty-eight hours of operation may be excluded in reckoning percentages. SYMPTOMATIC PAROT[TIS 37

ACUTE PAROTITIS.

1. PAROTI'TIIS FOLLOWING OPERATION ON AND DISEASES OF THE ABDOMINAL VISCERA.

Ga8tric O)peration8 and Disea8es. Of the various primary disorders that nay be followed by this complication lesions of the stomach would appear to furnish the greatest proportion of cases. I find that the complication reaches its highest percentage in connlec- tion with operations for gastric ulcer. During the fourteen years 1890 to 1903, 49 operations were performed in St. George's Hospital for gastric ulcer, 42 of which were for perforated ulcer and 7 for non-perforated, 5 of the latter involving opening of the stomach. Among the :38 patients who survived the operation for forty-eight hours there were 7 cases of parotitis, which gives a percentage of rather over 18 among surviving cases. The cases of perforated gastric ulcer formed the subject of a paper read before this Society last year by Mr. T. c. English (3). Parotitis formed only one of the numerous septic complications that arose in the surviving cases. Among 42 patients with perforated ulcer, of whom 32 survived operation for forty-eight hours, 15 developed pleurisy, 2 empyema, 3 pneumonia, 1 pulmonary abscess, 3 thrombosis, and 1 acute nephritis. Nearly all the patients that developed parotitis had some other septic complication. In case No. 1 of my series suppurative parotitis was followed by pulmonary abscess, empyema, and pericarditis. Cases Nos. 2 and 4 developed pleuLrisy, and No. 3 pneumonia, in addition to parotitis. These complications are to be ascribed to the septic soilinog of the peritoneum that occurs with the perforation of the ulcer. In cases of operation for non-perforated ulcer, other septic complications were, as might have been 38 SYMPTOMATIC PAROTITIS anticipated, less frequent. A numiiber of instances of parotitis after operation for gastric ulcer will be found in the literature. Dr. Phillips and Mr. Silcock (4) and Dr. Aitkeni (5) have recorded cases. In the former case the patient also developed a sub-diaphragmatic abscess and in the latter septic pneumonia. Dr. Blumer (6) records a case following excision of a gastric iulcer ; and Mir. Mansell Moullin, (7) in his report on thirteen cases of operation for recent gastric ulcer, mentions parotitis as occurring in two of the patients. One case in Morley's series followed operation for gastric ulcer. The relative frequency of parotitis after other gastric operations performed at St. George's Hospital is showin by the following table, which summarises the operations per- formed during a period of fourteen years: No. of No. of No. of cases surviving cases of Operation. cases. 48 hours. parotitis Gastro-enterostomny. 53 43 3 Gastrostoiny 35 31 1 Division of adhesions .7 7 1 Gastrotoiny 5 5 0 Gastroplication 2 2 0 Pyloroplasty .o 5 0 Loreta's operation 2 2 0 109 95 That is to say, parotitis occurred in rather over 5 per cent. of the patients that survived operation forty-eight hours. If we add to these the gastric ulcer cases, making in all 158 cases of operationi on the stomach, with 13:3 patients surviving forty-eight hours, and 12 cases of parotitis, we find a percentage of 9 for all gastric opera- tions. It will be noticed that the incidence of parotitis is considerably less in the cases contained in the above table (5 per cent.) than in cases of perforated gastric ulcer (18 per cent.), which is suggestive in view of the lesser degree of septic soiling of the peritoneum in the formier. Other septic complications are also less common than in cases of perforation. Case No. 12 in my series developed SYMPTOMlIATIC PAROTITIS 39 thiomlbosis in additioni to parotitis after the operatioln of gastro-enterostomny. Cases of parotitis following operations on the stomach xvill be founid in AIr. Paget's and in Alorley's series, and a case followiing gastrotomy for the remolval of a hair-ball is recorded by the late MIr. Knowsley Thornton (8). These figures would show that parotitis occurs as a well-marked complication of gastric operations, and that its frequency increases in proportion to the amount of peritoneal .

[L Cases of Gast-ic Ulcer not treated by Operation. Although much less frequent than in the operative cases, parotitis forms a vell-marked com-plication of cases of gastric ulcer treated by medical measures. It occurs with such regularity that it seems remarkable that it has beeli so little recognised as a complication of this disease. Taking 652 consecutive cases, extending over a period of thirteen years, I found that it occurred thirteen times, or in exactly 2 per cent. of the patients. There was never any suspicion of an epidemic among the patients. The fact that the average period of starvation or rectal alimeenta- tioni in these cases was so much longer than in the operative cases is certainly suggestive of the inadequacy of the usually accepted explanation that parotitis is due to the cessation of mouth feeding, especially when wve consider the relative frequency of the complicat'ion in the cases treated surgically and those treated medically. Miloreover, the infectioln was of a much milder type in these mnedical cases, for in two only out of fifteen was there any suppuration, and in one of these it involved only the lyvnphatic glanlds over the mastoid process, and not the main parotid. In one case (No. 44) thrombosis occurred in addition to parotitis. A peculiar case is recorded by Dr. Sydney Phillips (9). A girl, aged 19, came under observation with symptoms strongly suggestive of perforated gastric ulcer, but appar- 40 4SYMPTOMATIC PAROTITIS ently due to thrombosis of the inferior veena cava, whiclh proved fatal. The patient developed parotitis, and a few days later the thrombosis, spreading dowinwards, appeared in the veins of the leg. Extensive thrombosis, extending right through the iliacs and inferior cava, was found at the autopsy. The patient was on rectal feeding for one day only. A possible explanation would seem to be that the parotid was predisposed to infection-the patient had had dyspeptic symptoms for some time-and that the organisms which were responsible for the thrombosis, reaching the gland by the blood streanm, set up the inflammatory process.

Lesioqns of the Intestinal Trcact. Parotitis inay occur as an after-complication of opera- tions on the intestinal tract, but less commonly than after gastric operations. Its frequency seems to diminish as the site of operation passes dowvn the gtut. It is almost unknown after operations on the rectum; but cases following division of a rectal stricture and the excision of hemorrhoids have been recorded by Paget and Auld. A possible source of fallacy is to refer this complication to the operation rather than to the under- lying condition which the operation is intended to relieve. For instance, two of the cases in my series followed the operation of colotomy in patients with acute intestinal obstruction ; but inasmuch as I find that parotitis is a common complication of acute intestinal obstructioin, while we have had no instances of its following colotomy- performed for other conditions, I prefer to assign the parotitis to the intestinal obstruction rather than to the operation. One of my cases (No. 49) occurred in a patient with duodenal ulcer. The symptoms closely resembled those of enteric fever, and the patient was treated accordingly, and fed by the mouth. A large chronic non-perforated ulcer of the duodenum was discovered at autopsy. SYMPTOMATIC PAROTIT'1IS 41 AMr. Paget has recorded a case of parotitis in connection with perforated duodenal ulcer going on to general peritonitis and ending fatally. One would expect this comiiplicationi to be fairly frequent in cases of perforated duodenal ulcer. I have not met with any instances of it, but our records only contain eight cases in fourteen years, and of these three only survived operation for forty-eight hours. I have not been able to find any instances of this comil- plicatioln after operation on the small or large intestine in our records, except, as I have mentioned, after colotomy performed for acute obstruction. These operations are summarised in the following table: No. of cases Operation. N'o. of cases. surviving 48 hours. Intestinal resection, anastoniosis, etc., mostly for injury 30 16 Intussusception. . . . 22 10 Colotomy for other conditions than acute obstruction 140 . 133 192 159 Intestinial Obstruction. Three cases of parotitis in connection with intestiinal obstructioln will be found in Mir. Paget's series published in the 'Lancet' in 1886, in two of which enterostomy was performed. Morley also records cases after eniteros- tolmiy and operation for obstruction. I have found parotitis quite a commoln complication of acute intestinal obstruction, as the following table shows: Disease. No. of Surviving Cases of cases. 48 hours. parotitis. Acute obstruction due to bands, volvulus, impacted foreign bodies, etc. 50 31 3 Acute supervening on chronic obstruction due to malignant disease, tuberculous peritonitis, etc. 30 21 2 80 52 5 This gives a percenltage of nearly ten amolng surviving cases. 42 S42YMPTOMATIC PAROTITIS In one of the ovariotomy cases in my series (No. 28) the removal of the cyst was followed by acute intestinal obstruction, due to kinking of the gut, about the time that the parotitis developed. These cases of intestinal obstruction (Nos. 13 to 19 in my series) mostly bear clear evidence of intra-peritonieal sepsis. In No. 15, a case of volvulus, more than a foot of intestine was found intensely congested, and in places gangrenous; there was also much infiltration of the mesenterv. In No. 16 the obstruction was due to a band, and the gut was almost ulcerated through at the site of constriction. At the autopsy the injured gut was found to have given way, and general peritonitis was present. No. 17 was also a case of obstruction by a band; the patient recovered, but with free suppuration of the wound. In No. 19, where the obstruction was due to carcinoma of the sigmoid flexure, the gut was also found at the autopsy to have given way, and general peritonitis was present. In addition to parotitis the patient also developed a left-sided pleurisy. The frequency of parotitis as a complication of these cases of acute obstruction is quite explicable if we con- sider that we have not only the enormous amount of gastro-intestinal disturbance which accompanies these cases to dispose the gland to infection, but also the presence of intra-peritoneal sepsis to act as a primary focus of infection for the parotitis and other septic com- plications. Strangulation of external hernive is, of course, a variety of intestinal obstruction, and is occasionally followed by parotitis, but less frequently than the internal variety. The reason for this would appear to be that the condition is sooner recognised and treated, and also because it does not involve an abdominal section, with its necessary damage to the peritoneum and abdominal contents, which apparently acts as a strongly disposing cause to the de- velopment of this comnplication. I found four cases of parotitis among 346 cases of strangulated hernia, 303 SYMPTOMNIATIC PAROTITIS 43 patients surviving operationi for forty-eight hours; that is to say, it occurred in rather over 1 per cent. of cases. These four cases (Nos. 20 to 23) also illustrate the association of parotitis with other septic complications. No. 20 developed throinbosis, No. 21 pleurisy with effusion. In No. 22 the parotitis was accompanied by pneumonia, and four days after the onset of these com- plications the patient suddenly collapsed and died, apparently from embolism. In No. 23 there was a very mild attack of symmetrical parotitis lasting only thirty-six hours, and no other complications. Mr. Paget collected several cases of parotitis followinig herniotomy and Morley's series also includes a case. Mr. F. B. Jefferies (10) records a case where it followed operation for strangulated umbilical hernia. The parotitis was associated with bronchitis, and the tissues of the nieck became so swollen and infiltrated that there was imarked dyspnleea aind dysphagia, which apparently largely determined the fatal result. I have found only one case of parotitis after operation for noln-strangulated herinia (Case No. 24). In this case there was a large ventral hernia, and the operation was practically an abdominal section. Operation was followed by much vomliting, and the wound suppurated freely. In addition to suppurative parotitis there was an abscess of the soft . There have been no instances of parotitis amonig 8;5 cases of simple " radical cure."

Ulceration of Intestinle. Just as parotitis m-lay comnplicate gastric oIr duodenal ulcer, so it mnay occur with ulceration lower down the gut. Case No. 51 was an instance of ulcerative colitis, the ulceration being found after death to be confined to the coecum anid beginiing of the ascending colon. The patient was admitted with h'miateimiesis and nmelelna and was fed for somiie timie by rectumui. In Case No. 52 the patient 44 SYMPTOMATIC PAROTilIS had extensive follicular ulcerationi of the large, and to a less extent, of the small intestine. T'he symiiptomlis in this case were intractable diarrhoea and vomiiiting. An early Stage of peritonitis was present at the timie of death. The patient was fed entirely by mouth. Case No. 57 occurred in connection with a severe attack of gastro- enteritis in an infant of four months. Foul days after the onset of symptoms the left parotid became inlvolved, and two days later again the right. Both parotids suppurated, as well as some of the cervical glands. A brother of this patient was said to have had an exactly simiilar association of suppurative parotitis with gastro- enteritis which proved fatal. In this connection-the comparatively frequent associa- tion of parotitis with enteric fever is interesting.

Parotitis int Ap.)pendicitis. Parotitis mnay occur in appendicitis, but is almiiost always associated with intra-peritoneal 'sepsis, either in the formll of abscess or peritonitis. It is apparentlyinot particularly common, for there was no instance among the 515 cases in our records during the years 1890 to 1903. Quite recently, however, a case of parotitis has occurred in a patient in the hospital, following appendicectomy (Case No. 27). The patient, a girl aged 16, had suffered for some considerable time from pelvic pain. Laparotomy being performed, the pelvic organs were found to be apparently normal, as was also the appendix, which was reemoved. Left-sided parotitis came on five days after operation, with dryness of mouth and some febrile dis- turbance. It subsided within a week; the right parotid was not affected. The occurrence of parotitis, without appendix abscess and the affection of the left parotid alone, are somewhat unusual features. The first recorded case of parotitis comiiplicating appendicitis is apparently Barlow's, whicli waas published in 1886. Of Morley's cases, two followed "removal of SYMPTO0MAT'IC PAROTITIS 45 the appendix." Mr. Paget also records cases; a referenice to his series in the ' Lancet' of 1886 xvill illustrate the class of case in which this compllication occurs. Thus, inl Case No. 14 of his series there was general peritonitis; in No. 15 the autopsy revealed peritoneal abscesses, some of great size, perforation of the appendix, and signs of old genieral peritonitis; in No. 19 general peritonitis and appendix abscess xvere present. Other cases are recorded by Stanley Thomas (11), Elder (12), Barjon (13), and Fiske Jones (14). Appendix abscess is definitely stated to have been present in all except the first of these. Fiske Jones' case is interesting, because the parotitis recurred with recurrent attacks of appendix abscess-there were three attacks in all. The attacks of paiotitis always followed exactly the same sequence. Exactly forty-eight hours after each operation the right parotid became swollen, and forty-eight hours later again the left. The parotitis took six or seven days to subside. A very interesting case has recently been published by Mr. Brennan Dyball (15), in which the parotid infection was extraordinarily severe, and is considered by the author to have been the imniediate cause of death. The appenidi- citis was of the fulhuinating type, and at the operation, performed shortly after the onset of syimptomiis, the appendix was found to be gangrenous, with acute in- flammiiation of the meso-appendix and free pus in the peritoneal cavity. The parotid sloughed in this case, although early incisions were made. The teeth in this patient were in a state of advanced caries, and the author attributes the parotid infection to the septic oral cavity. The patient, however, exhibited symptoms of profound toxainiia or septicsemia very shortly after the implication of the parotid and long before the parotitis had reached its height. The association of a virulent infection of the appendix, going on to actual gangrene in a couple of days, with an acute infection of the parotid, which also ended in gangrene, is at least suggestiv-e of the origin of the parotitis in the abdominal infection. 46t SYMPTOMATlC PAROTITIS

F. E. Bunts (38) has recently published a paper oII "Parotitis complicating Appendicitis." After a refereince to some of the above cases, he describes three cases that came under his own observation, and a fourth under the care of Hamann. All four were presuilmably cases of abscess, although two only came to operation. The parotids suppurated in all four cases. In two of the cases the pus both from the appendix and parotid abscesses was exanmined bacteriologically. The organisimi isolated from the appendix abscess was the Bacillus coli comninunts in both cases-the parotid abscess in one case furnislhed the coli comi-munis and the Staphylococcus pyogenes aureus; in the other, the stapliylococcus only. Bunts lays mliuch stress on the examination in this latter case-where differelnt organisms were found in the two abscesses-as tending to show that the parotid infection could not have arisen from the abdominal abscess. It is conceivable, however, that a staphylococcus may have been present at an earlier stage in the appendicular lesion, aind have been subsequently outgrown by the coli communis. It is interesting to notice that in cases of appendicitis there is a greater tendency for the right parotid to be affected than the left. If the affectioni is bilateral, it usually starts in the right parotid.

Parotitis following Abdominial Operationts on the Pelvic Organis. I shall consider this class of operation at the present time, because a careful consideration of the cases would seem to point to the damage to the intestines anid peritoneum, involved in these operations, and peritoneal sepsis as being more important factors in the causation of parotitis than some obscure sympathetic connection between the organs of generationi and the salivary glands. The cases of parotitis following operations on the generative organs in which the peritoneal cavity is not directly involved I shall consider later. SYMPTOMATIC PAROTITLS 47 In Mr. Paget's series of 101 cases, somiewhat over one half followed interference with the pelvic organis, twenty- seven cases alone following ovariotomiiy. Of fifty-one cases of post-operative parotitis collected by Morley in America, twenty-six followed ovariotomy and eight more abdominal operations on the uterus and its appendages. Benoit's thirty-five cases, forming the subject of a Paris thesis, were all collected from the records of ovariotomy. Moeericke asserted that he had never seen it except after this operation. It is apparently the preponiderance of cases following the operation of ovariotomy in these series that has led to the statement, found in most text-books, that ovariotomy is the operation which is most prone to be followed by the complication of parotitis. This is perhaps to be explained by their collection from the records of a time when ovariotomy was the only abdominal operation extensively practised. But an examination of series of consecutive cases will show that the incidence of parotitis after abdominal operations on the pelvic organs is relatively low. Our obstetric in-patient records for twenty-one consecutive years show that among 3065 ini-patients, with 400 laparotomies, there were six cases of parotitis only. The cases are distributed as follows: Cases Operatioii No. of Cases. of Parotitis. Ovariotomy . . .225 4 Operations on Fallopian tubes 37 1 Rupt. ectopic gestation 3. 1 Other abdominal sections 105 0 400 .6 Parotitis, then, would seeimi to occur in from onie or two per cent. of cases only; it is about as frequent as in cases of exploratory laparotoilmy, and much less commiion thani after gastric operations. This complicationi was relatively more frequent in earlier days. In the records of 300 cases of ovariotoimy published by Kniowsley Thorntoni (16) parotitis is imentioned as occurring in three 48 SYMPTOM3ATIC PAROTITIS cases. Moeiricke recorded five cases in 200 ovariotoiimies, Goodell four in 153. Of the cases in my series (Nos. 28 to 34) four followed remnoval of ovarian cysts, one exploratory laparotomV and tapping of a malignant cyst, one an operation for ruptured ectopic gestation, and one removal of the appendages for hydrosalpinx. The points that these cases emphasise are -the accompanying damage to the abdominal contents involved in these operations, and the presence of peritoneal sepsis. In the case of ruptured ectopic gestation (No. 28), the placenta was left in situ, the umbilical cord brought out of the abdomen, and the placental site drained. Peritonitis supervened a month after operation, and three weeks later, again, parotitis. Subsequently symptoms of obstruction came on, and the patient died. At the autopsy genieral peritonitis was found, with an abscess at the placental site. The parotitis in this case is clearly to be referred to the peritonitis, and not to the original operation. A similar case after operation for ruptured ectopic gestation is recorded by Lewers (17). In his case the abdominal wound reopened spontaneously and a purulent discharge set in. The same points are showin in the ovariotomny cases. In Case No. 29, shortly after the onset of parotitis, the patient exhibited syinptomns of acute intestinal obstructiol. On reopening the abdomen a coil of intestine was found kinked by a band; there was also some peritonitis present. In case No. 30 the cyst was adherent to the intestines, and the adhesions had to be broken down. The patient died of general peritonitis, which came on at the timne the parotitis developed. In No. 31 the cyst was so adherent that it was left and drained. The case was also coini- plicated by tuberculous salpingitis and peritonitis. The discharge from the wound rapidly became purulent, and the symptoms were generally septic. Parotitis developed a day or two before the patient's death, three weeks after the original operation; it is clearly to be assigned to the SYMPTOMATIC PAROTITIS 4'3 peritonitis and niot the operation. In No. 32, again, there were many adhesions, the division of which necessitated much damage to the intestines. The patient in this case, in addition to parotitis, developed not only pelvic peri- tonitis, but also symptoms of general pyaemia. In No. 33, where the appendages were removed for hydrosalpinx, there were many adhesions. In No. 34 a mialignant ovarian cyst was found at operation and tapped; there were many secondary growths about the peritoneum. Parotitis developed three weeks after the operation, shortly before the patient's death, and probably other factors than the operation were responsible for it. The other recorded cases of parotitis after operation on the pelvic organs also bear clear evidence of sepsis. Of the three cases in Knowsley Thorntoni's series, one died on the tenth day of pyemia, and another on the ninth day of septicmnia. An almost silmlilar frequency of the terms " pyaemia" and "peritonitis" will be found in the ovariotoilly cases recorded in Mr. Paget's series in the 'Lancet' of 1886. Of the seven cases in my series following operations on the pelvic organs the parotitis followed the operation sufficiently closely to be referred to it in four cases only. In two cases it developed only with the onset of septic peritonitis. In one case it was associated with acute intestinal obstruction. Int most of the cases there is evidence of damage to the peritoneum and intestines at the time of operation, and in all but the sixth and seventh of intra-peritoneal sepsis. The proportion of cases, then, that can be referred unequivocally to damage to the pelvic organs dwindles still more, and one is justified in saying that on the whole parotitis is not a very common compli- cationi of operations on these organs. Since writing this I have come across an observation of Brunn's to the effect that cases of parotitis after ovariotoiny were most frequent when multiple adhesions between the viscera were en- countered, which bears out my own observation. One case of parotitis (No. 51) occurred in connection with pelvic peritonitis, treated without operation. 50 SYMPTOMATIC PAROTITIS

Parotitis after other Abdomninal Operation,s. There is practically no operation which involves the opening of the peritoneal cavity which may not be followed by parotitis. Among 265 cases of abdominal section, including operatioins on the liver and gall-bladder, spleen, pancreas, and kidney-operations for tuberculous peritonitis, injuries, and new growths-exploratory lapar- otomies, etc., I have found two cases of parotitis. In both cases (Nos. 25, 26), one of cholelithiasis, and the other of tuberculous peritonitis, the operation was siimiply exploratory.

Parotit is int connection with Diseases of the Liver anzd Gall- bladder. Three of the cases in miiy series occurred in coinnectioni with cholelithiasis-in one case (No. 26) following an exploratory operation. In all three patients there was a great deal of gastro-intestinal disturbance in the formii of voimiiting, and in one case, in addition, diarrhoea. One patient (Case No. 26) was admitted with symiiptoms suggestive of a mild degree of in-testinal obstruction. The other two cases (Nos. 54 and 55), which were treated without operation, had had violent attacks of vomiting and colicky pain. The mouth in all three patients was extremely dry on admission. In case No. 55, in which the parotitis went on to suppuration, pyeemic abscesses subsequently developed elsewl-here. The feeding in these patients was by the mouth. The gastro-intestinal disturbance in these cases is probably the cause that disposes the gland to inifection. In all three cases the onset of parotitis was preceded for some considerable time by suppression of secretion. In these cases there is probably a primary septic focus within the abdomen. It is well kn-owi-n that in cholelithiasis the infection of the gall-bladder may spread outwards, giving SYMPTOMATIC PAROTITIS5 51 rise to a localised peritonitis, which, as Mayo Robsoni has pointed out, may be sufficiently severe to cause intestinal obstruction, either by giving rise to adhesions or by paralysis of the intestine from peritonitis. A certain degree of intestinal obstruction was exhibited by the first of my cases. In one of the patients the parotid abscess was merely one of several pywimic abscesses; in such a case the path of infection is almost certainly the blood- streamii, and the primiiary focus the septic gall-bladder or the inifected tissues in its neighbourhood.

The Connection between Parotitis andt Diseases of the Pancreas. The pancreas and the salivary glands are so similar structurally and physiologically that coincident involve- menit of the two organs would not be at all surprising. It has been observed that in certain diseased conditions of the pancreas the salivary glands mlay be affected, iiostly in the direction of alteration of their secretion-as evidenced by the term " pancreatic salivation." Mayo Robsoln (18) has recently recorded a case of parotitis in a patient with suppurative catarrh of the paniereatic ducts associated with suppurative cholangitis. Right-sided parotitis was present, and the parotid was incised at the time of operation. According to Robson, suppurative catarrh of the panicreatic ducts is generally, if not always, comiibined with suppurative cholangitis, and inasmuch as parotitis is not ani infrequent complication of cholelithiasis it is difficult to say which condition was responsible for this comiiplication. One would imaginie that the pancreatic condition would more strongly dispose to the development of parotitis than the hepatic, and it is possible that the pancreas might be involved in these long-standing cases of cholelithiasis that develop parotitis. I know of no published cases of parotitis in uncomplicated pancreatic disease, but the study of these diseases is comparatively recent. Apparently the opposite condition of secondary VOL. LXXXVIII. 5 .52 SYMPTOMATIC PAROTITIS involveiienit of the pancreas in inflammiiatory conditions of the parotid does obtain-at all events in the case of epidemic . Cuche (19), Priestley (20), Jacob (21), and Simonin (22) have all described pancreatitis as a complication of ordinary mumps. Simonin found symptoms of acute pancreatitis in ten cases out of 652 patients in an epidemic of mumps, or in 1-3 per cent. Whether the pancreas is affected in cases of symptomatic parotitis or not I have no knowledge. I have examined the post- mrortem records of cases dying after the development of parotitis, with especial reference to the condition of the pancreas, but these are not sufficiently extensive to throw much light oni the subject. The condition of the pancreas is inentioned in seven- of these cases. In four it is described as being normnal, but in three of these the parotitis had commenced fifteen, thirty-two, and forty-one days respectively before death. In three cases there was evident naked-eye alteration in the gland. In case No. 53, a patient who died with follicular ulceration of the intestine and who developed parotitis a fortniight before death, the pancreas was found at autopsy to be congested; in case No. 61, a case of lobar pneumonia, in which parotitis developed five days before death, in conjunctioni with multiple abscesses-a condition of interstitial pan- creatitis was founld at the autopsy; and in case No. 67, a patient who died of erysipelas and in whom all the salivary glands were inflamed and swollen, the pancreas was found to show fatty infiltration. These results are rather hetero- geneous, but I think that it is quite possible that with increasing recognition of pancreatic symptoms the pancreas may be found to be involved in a certain proportion of these cases-as it is in epidemic mumps. I shall have to refer later on to parotitis in commectiomi with gout. It is interesting to note that Portal in 1804 recorded a case of acute suppurative pancreatitis followimig an attack of gout in the feet. Dr. Peacocke has described a case of parotitis in a patient suffering from (23). The patient was admitted to SYMPTOMATIC PAROTITIS 5:3 hospital in a conditioii almost of coma, an-d died a imionth later. The parotitis developed about a fortnight after admission and was accompanied by slight fever, which con- tinued till death. It appears possible that the parotitis in such a case as this might be due to some underlyinig disease of the pancreas, at any rate as a predisposing cause.

Parotit8 iit connitection with Peritonitis. Trhe association of parotitis with peritonitis in abdomiinal cases is so close that one would lhave no hesitation in say- ing that peritonitis is the commonest immediate cause of parotitis. Although it is convenient to classify the cases according to the primary condition present, such as perforated gastric ulcer, appendicitis, and so on, it inust be remembered that the imiajority of cases are to be referred to peritoneal infection in the first instance. In imiany of my cases the developmi-ent of parotitis coincided in point of time, not with the operation, btut with the onset of peritonitis. In series of cases taken from earlier days, when exploratory laparotoiny was less common, many instanices aire described in connection with acute peritonitis. IMr. Paget's series cointainis eight instailces. Addenbrooke (37) has described a case in connection with acute suppurative peritoilitis in a boy ; Hutchinson (40) a case in coinnection with acute peritonitis following rupture of the duiodenum. Case No. 51 in mny series occurred in connection with pelvic peritoniitis. In many of the earlier cases the acute perntonitis was secondary to appendicitis. Ont the Catnsation of' Parotitisfrloowing Operationts ont antd Diseases oj the Abdominial Viscera. Several theories have been advanced from timie to time to account for this complication. In his paper, published in 1887, AIr. Paget discusses the various hypotheses that had been formulated up to that time. He discarded firstly the idea that it was a pyaemic or septiceemic manifestation, because in 93 per cent. of the cases in his series parotitis 5O'4 SYMPTO3ATIC PAROTI'TIS was a solitarv comliplicatioin. The fact of a comiiplication being a solitary one does not, however, preclude the possibility of hEemiiic infectioni. Secondly, he concluded that it was not due to "iinflaniniation extending froim a parched anid sordid iiiouth, up to the duct of the gland," because on this view the socia parotidis should be the first part affected, which did not appear to be the case. He concluded that the reflex action of the nervous system played a large part in its causation, instancing the physiological experiment that interference with the alimen- tary canal of the dog hinders or arrests salivary secretion. An attack of acute parotitis. is, however, something more than a mere suppression or retention of the secretion. The viewimost generally held is that in the absence of food from the mouth the salivary flow is not stimulated; the mouth consequently becomiies dry and septic, and an opportunity is afforded for a spread of infection to the gland by the duct. AMartin has endeavoured to attribute the dryness of mouth to the effect of an2esthetics. The fact that parotitis is by no means confined to patients who have undergone anisthetisation shows this view to be inadequiate. Of the Germanl observers, Pfannensteil tr aces it to infection from the mouth, whilst Moericke, arguing from gyiiwcological cases, puts it down to a sympathetic relation between the ovaries and salivary glands. Such a relation, if it exists, cannot do more than dispose the gland to infection. There would appear to be a better marked nervous relation between the salivary glands and the gastro-ilntestinal tract. Benoit, in his thesis, concludes that parotitis following ovariotomy is usually due to a spread of infection from the mi-outh, but the infection miiay occur by the blood- stream. Condamin, who criticises Benoit's conclusions, is very averse to the idea of a haiemic infection. Morley, whose series is taken largely fromii gyniicological cases, lays stress on the dryness of mouth following abdominal operations on the pelvic organs, which lie SYMP'TOMATIC PAROT'I'IS ascribes to a symiipathetic relation betweeni the gelnerative organs and the salivary glands. The actual inifection he conisiders occurs fromii the mouth. Pozzi considers that the onset of parotitis after coeli- otomiiy is evidence of a certain degree of septicaeiiiia. Osler, like Benoit, thinks that infection may occur both by the duct and the blood-stream. The mnethod of infection is discussed at lengtlh by Bunts in hiis recent paper (loc. cit.) ; he argues for infection by the duct, stating that in-flammation of the duct has in somlze cases been found to precede inflaiimmation. of the glalad. Injutry to the gland or duct during operation or by the ain-esthetic, which he suggests as a contributory cause, nust play a very minor part, because there is no reason why this should occur especially in abdominal operationis, to which the complication is practically confined. Nor, as I have shown, is this form of parotitis necessarily post- operative. After arguiiig stron-gly in. favour of infection by the duct, Bunts nevertheless, at the end of his paper, admits the possibility of a h,emic infection. To quote his words: " Duct infection, in the light of modern pathology, anid from the clinical study of maniy cases, would seein the only tenable theory; and while no doubt a large percentage of these infections occur primarily in Stenson'is duct, the knownv fact that many patlhogenic are cast off through the salivary glands makes it not im- probable that, the locts rminoris resistentidi having been prepared by the dryness of the mouth, the stagnationi of the saliva, the iinactivity of the imiasticatory muscles from the enforced liquid diet or from possible injury to the duct or gland in operation, the bacteria may develop directly in. the parotid, eveni while the duct is the site of an in- flammation derived from bacteria in the buccal cavity." The various observers, then, for the m-lost part recognise a predisposing cause-the suppression of secretion, which is variously attributed to reflex niervous effects and to the absence of food froimi the mnouth, and an excitino cause, infection. The actual iniflammilnation is always preceded by 56 SSYMPTOMATIC I'AROTITIS suppression of secretioin-this is conistanlt; the parotitis itself is inflamlmnlatory and septic. The process presenits all the appearances of ain acute iniflamimiation; in severe cases the attack may even go on to extensive cellulitis of the surrounding structures, involviing in some cases the whole of the side of the face and neck. It is allmost inivariably accompanied by febrile disturbance-tlhere was a rise of at least 20 in temperature in 34 out of 44 cases in my series that followed abdominal lesions. The frequent terminationi of the process in suppuration is again coii- clusive of its nature. The older view that the swelling of the gland might be merely passive and due to a retention of secretion is quite inadequate. We have, therefore, to deal with two stages in the process. The first stage consists of the suppression of the secretion, a process which throws the gland out of its normal condition, and renders it more liable to infectioni, and the causes that operate to produce this suppression of secretion may be called the predisposing causes. The second stage consists of the actual , and the cause at work in this process-the exciting cause-is infection. These may be considered separately.

A. The PIredisposing Causes leadingg to SuppressioiI of Secretioni. That the onset of symptomatic parotitis is preceded by suppression of the salivary secretion is an old observ-a- tion. 'rhis suppression has been asciibed to the abseince of food from the mouth ; it is suggested that in default of their natural stimuli the saliv-ary glaiids are not excited and cease to secrete. But, although the abseniee of food nmay assist in the suppression of the secretioni, it is prob- ably not the main cause. In my series of cases not only had mlany of the patients been receiving solid food by the mouth for considerable periods before the onset of parotitis, but in many cases-in fact, in over one-third- imoutl-feedinig had never been interrupted. Even amonig SYM3PTOMATIC PAROTITIS 51, the fifteen iion-operative cases of gastric ulcer, six onily were on rectal alimentatioln at the time of the onset of parotitis, and two of the cases had been on miiouth-feeding for ten and eleven days respectively. We must therefore look to other causes for the suppressionl of the secretion. These causes are to be found in injury or disease of the alimentary canal. Apart from the direct experiment quoted in Mr. Paget's paper that interference with the alimentary canal of the dog leads to a suppression of salivary secretion, we have abundanit evidence clinically of the effect of abdominal operations on the secretion. Thirst and dryness of mouth are always more marked after abdominal than other operations, and I have always found this to be especially manifested after operations on the stomach. Dryness of mouth is also especially notice- able in cases of peritonitis. Injury or disease of the gastro-intestinal tract seems to have the effect in some way of inhibiting salivary secretion. This samne inhibition may also be marked after operations oni the pelvic organs, and I would suggest that it may be due not so much to the interference with the pelvic organs themselves as to accomiipanying damage to the gastro-intestinal tract or peritolieum. The probable cause of this suppression of secretion is a reflex niervous inhibition, set up by reflex nervous impulses from the injured alimentary canal. There is a possibility that another cause may be at work as well. Bayliss and Starling have pointed out the qo4le played by the body "secretin" as an excitant of pan- creatic secretion. Enriquez and Hallion (24) state that Lamibert and Meyer find that the presence of this body in the circulation accelerates the salivary secretion, although this is denied by Bayliss and Starling (25). It is obvious that grave lesions of the stomach, leading to a temiiporary paralysis of that organi, or such a condition as l)eritonitis, would largely interfere with the formation of this body, anid might in that way lead to a dimiinution of the salivary secretion. In the absence, however, of more exact knowledge on this point, we must allot the chief 58 SYIYMPT1OMA'TIC PAROTITIS share in the suppressioni of the salivary secretioln to reflex nervous influelnces. rThe fact, too, that, with a unilateral primary lesion, the parotid of the samne side seems mliore pronie to become affected also poinlts in the same direction. Suppression of the secretion of the gland, in whatever way caused, by throwing the glanid out of play, would seem to lower its resistance and leave it prone to infection. It remailns, therefore, to consider the method in which the gland becomes infected. The exciting cause, infection.-The direct or exciting cause of parotitis is ulndoubtedly ilnfectioln; the only question is the path by which it reaches the gland. There are three possible ways in whicn infection might gain access to the glanid-firstly, by the blood-stream, secondly, from ti mouth along the duct, and thirdly, by direct spread of an inflammatory process from surrounding structures. The more genierally accepted explanation is that infection takes place by the duct ; Osler, Benoit, Pozzi anid Bunts, however, consider that it may occur by the blood-stream. My own opinion is decidedly in favour of the latter view, and I will endeavour to state as briefly as possible my reasons for this view. 1. In the first place, if we examine carefully the cases of parotitis that occur after abdominal operations, we shall find abundant evidelnce of a primary septic focus within the abdomen. In the thirty-four post-operative cases in my series there was actual peritonieal sepsis in nineteen and suspicionl of it in others. I have drawn especial attentioni to this in the preceding part of this paper. An anialysis of the sixty cases collected by -Mr. Paget and published in the 'Lancet' for 1886 demonstrates clearly the importance of this septic element.

Cases. Evidences of sepsis. Five cases following use of cath- Chronic cystitis; operation fol- eter or sound lowed by rigors, pyrexia, etc. Case following introduction of Operation followed by rigor and pessary temperature of 104-5°. SYMPTOMATIC PARIOTlTIS 59

(Cases-cont. Evi(leluces of sepsis-cost. Seven cases after delivery or Pyamiia, two cases; peritonitis, abortion two cases; continued fever, one case. Three cases of appendicitis General peritonitis in all. Pyo-salpinx Pyaemia. Bullet wound, abdomen Pyaemia. Perforated duodenal ulcer General peritonitis. Gastric ulcer Vaginitis; bed-sores. Herniotomy Peritonitis. Remnoval of oimientum Pyaemia. Gono1aThsea Pyaeiiiia. Fifteen cases of ovariotomily Peritonitis, two cases; periniietritis, one case; pyameia, one case; death with septic symiiptoms, one case. Pelvic cellulitis Pelvic abscess and peritonitis So that there is direct evidence of sepsis in about one half of the sixty cases. If we turn now to the operations and diseases which I have found to be especially liable to this complication, the influence of intra-abdominal sepsis cannot be overlooked. I fincd that parotitis follows operationi for gastric ulcer in 18 per cent. of cases; whereas in cases of gastric ulcer treated without operation it occurs in but 2 per cent. And the operative cases are on the average confined to rectal alimentation for a shorter period thani the non- operative. One cannot but conclude that the septic soiling of the peritoneum in the case of a patient with perforated gastric ulcer has much to do with the pro- duction of this complication. After other gastric operations, such as gastrostomy or gastro-enterostomy, the percentage of parotitis is much lower, because there is less peritoneal contamination; but it is probable that in a certain proportion of these cases some slight con- tai-nination of the periton-eum does occur. I have pointed out that in practically all the cases of appendicitis in which parotitis occurred, which are recorded in detail, either abscess or general peritonitis was present. In cases of acute intestinal obstruction and stranigulated 60 SYMPTOMATIC PAROTI'TlS hernia, whlere damiaged gut is returned into the abdomiieni, there is certain to be some amiiount of-it may be localised -peritonitis. In somlie of the cases in miiy series the daimaged gut was found to have given way, and a condition of general peritonitis to be present. In the gyniicological cases, again, there was ablundaiit evidence of peritoneal sepsis. 2. The second point is the frequent associationi of parotitis with other septic complications. I have already mentioned the large proportion of septic complications that Mr. English has found to complicate the after course of patients with perforated gastric ulcer. All our cases of perforated ulcer that developed parotitis developed some other complication. Apart froimi peritonitis and intra- abdominal complications, and suppuration of the wound, I find that 11 of the 34 post-operative cases in my series developed other complications, such as thrombosis, pleurisy, secondary pneumonia, and metastatic abscesses. Other nmetastatic complications are, moreover, especially pronie to occur in patients who develop parotitis durilng the height of the specific fevers. When parotitis forimis olle of many secondary infections, it seells hardly reasonable to suppose that the inlfectioni in the case of the other comiiplications occurs by the blood, while in the case of parotitis it occurs by the duct. Nor does it seemii justifi- able to separate out the cases of parotitis whiclh occur with other septic colmiplications and label them py'eeimic anid adopt a different explaniationi for the parotitis that occurs as a solitary ev-ent. 3. A careful study of the cases, imoreover, seem-iis to show that the severity of the parotid infection varies with the severity of the primary conidition and with that of the other complications, suggesting that the samiie organlisms are at work in both processes. Where parotitis occurred with other complications in the cases in mny series, if the other complicationi resolved, the parotitis usually did the samiie, while if the other comiiplicationi were a suppurati-e onie, the parotitis usually went on to suppuration. To SYMPTO31ATIC PAROTITIS 61 take an example, a resolving parotitis would be associated with a simple pleurisy, a suppurative parotitis with an eimpyema. The following table shows an analysis of the cases in which parotitis was associated with other compli- cations. Cases wher e Parotitis resolved.

ICase No. Disease or operation. Other complications.

2 Perforated gastric ulcer Pleurisy. 4

3 ,, Pneumonia. 12 Gastro-enterostomly Thrombosis. 44 Gastric ulcer 19 Intestinal obstruction Pleurisy. 20 Strangulated hernia Thrombosis. 21 ,, Pleurisy. 22 Pneuilmonia.

Cases where Parotitis vemt onz to Abscess.

Case No. Disease or operatioin. Other complications.

1 Perforated gastric ulcer Empyema, multiple pulmonary abscesses. 17 Intestinal obstruction Much suppuration of abdominal wound. 24 Ventral hernia Abscess of palate, sepsis of wound., 32 Ovariotomy Pyemia. 54 Cholelithiasis Multiple abscesses. 76 Remioval of exostosis I- The same point will usually be found to obtain in cases of parotitis occurring in the specific fevers-that a 62 SYMPTOMATIC PAROTITIS suppurative parotitis is associated with other ilnetastatic abscesses. MIoreover there is a genieral, though not absolute, rela- tion between the severity of the parotitis and that of the primary condition. In fifteen cases of gastric ulcer treated without operation, the gland itself suppurated in only one instance, whereas in cases where there was well-marled intra-peritoneal sepsis suppuration was much more fre- quent. Moreover in Mr. Dyball's case, that I have already quoted, of fulminating appenidicitis ending in gangrene of the appendix within a couple of days, and associated with symptoms of profound toxemia, the infection of the parotid was so intense that the gland actually sloughed. 4. The complication of parotitis was, relatively to the number of abdominal operations performed, far more frequent fifteen or twenty years ago thain at the present time. I find that there were almost as many cases of parotitis durinig the years 1890 to 1897 as there were during the years 1897 to 1904, although three or four times the number of abdominal sections were performed during the latter period. There appears to be some connection between the relative decrease in the niumber of cases of parotitis, and improvements in antiseptic surgery. If this coinplication were due to a spread of infection from the imouth, improvements in antiseptic surgery should have no effect on it. McDonald (39) has drawn atten- tion to its relative frequency in the early history of abdo- minal surgery. 5. As MIr. Paget pointed out, if the infectioll were an ascenidinig one by the duct, the first portion of the gland to be involved should be the socia parotidis. I have always looked for this involvement of the socia, and have found it in three cases, but it was always subsequent to the involvement of the main gland. The same process occurrediin Mr. Jefferies' case (loc. cit.). 6. With a view to testing the influence of oral sepsis, I have tried to observe as far as possible the condition of SYMPTOMATIC PAROTITIS 6.3 the mouth in cases that developed or seemiied likely to develop this complication. I could not establish any rela- tion between this complication and oral sepsis. Parotitis often occurred in patients in whom there was nio indication of a septic condition of the mouth or teeth, while patients with the foulest mouths escaped it. Moreover with a view to testing the efficiency of oral antiseptics, I analysed 300 consecutive cases of gastric ulcer. Among 68 cases in which antiseptic mnouth-washes were systematically employed, there were four cases of parotitis, while among 232 cases treated without thein there were only 2 cases; and in order to exclude other factors, taking the cases that were being fed by rectum only, parotitis was still four times as frequent in those in whom mouth-washes were used. Nor can it be objected that mouth-washes were used in patients with particularly septic mouths, for it is the practice of some to uise them as a matter of routine in all cases, while others do not. The only coniclusioni is that oral antisepsis has little effect in preventing the complication. To summarise the argument, theni, with regard to the path of infection. The tendency to the development of parotitis after abdominal operations seems to bear a marked relation to the amount of intra-abdominal sepsis. Further, the parotid infection, although it may be a solitary event, is frequently associated with other metastatic complications of haemic origin, and its severity varies with the severity of these and of the intra-abdomninal sepsis. Like other septic comiplications, its frequency has dimninished with improvemiients in antiseptic surgery. Lastly, the tendency to parotitis does not vary with the degree of oral sepsis, and measures adapted to combat this have not the slightest effect in preventing the complication. We may explain the process as follows: In these patients that develop parotitis there is a species of mild septicaemia. The organisms may be absorbed from the pe-ritoneal cavity, or, in the case of gastric operations or ulceration of the gastro-intestinal tract, through some 64 SYMPTOMATIC PAROTITIS of the rnucous surface, and-& reach the parotid by the blood-streamn. [n the case of a healthy functional gland, they might pass through it, as they pass through the kidney, and be thrown off in the secretion. But in these cases the secretion has been, reflexly or otherwise, suppressed, and the organisms, instead of passing through the gland, are retained within it and start an inflammatory process. The suppression of secretion acts, not only by lowering the resistance of the gland, but also by allowing the organisms to obtain a footing. In fact, in this pro- duction of parotitis after an abdominal operation, we have an exact parallel in those cases of acute nephritis follow- ing operation on the lower urinary passages, where the distribution of the renal lesions shows the path of infection to have been by the blood-stream. ln each case the effect of the operation is to depress the vital activity of the gland, and allow the organisms which would otherwise pass through it and be excreted to obtain a footing and start an inflammatory process. The only direct evidence for the theory of a spread of infection by the duct is an inflammation of the duct preceding that of the gland, which is said to have been observed in some cases. .Finally, if infection occurred by the duct, parotitis should be a common and not a rare phenomenon, since the necessary factors for such an infection are constantly pre- sent both in post-operative and in febrile conditions.

II. PAROTITIS FOLLOWING OPERATION ON AND DISTURBANCES OF THE URO-GENITAL ORGANS. I would divide these cases into the following groups: Parotitis in connection with-(1) Abdominal operations on the pelvic organs; (2) other operations on and diseases of the uro-genital organs; (3) parturition and abortion; (4) pregnancy and menstrual disorders. There is no doubt that disturbances of the generative organs mav be accompanied by disturbances of the salivary glands, and even by parotitis. In the camel the salivary SYMPTOMIATIC PAROTITIS 65 glands are said to eenlarge duriing the breeding seasoin. Clinically we are familiar with the metastatic orchitis, and the less common mastitis, vulvo-vaginitis, and ovaritis in the female, that may accompany ordinary epidemic mumnps. But a parotitis following a lesioni of the genital organs is not common, anid, except in Group 4, is confined practically to septic cases, and is probably of a pyaemic nature. The nearest parallel, perhaps, to the metastatic orchitis of munps are the cases of parotitis following injury to the testicle. Billroth recorded such a case following a blow on the testicle, and Hutchinson (40) states that Mackenzie com- municated to him two similar cases, both in young men. Group (1) I have already considered, and I have shown that inost of such cases bear evidence of intra-peritoneal sepsis, and that the accompanying damage to the alimen- tary canal alnd peritoneum may possibly act as the dis- posing cause of the parotitis. With regard to Group (2), illustrative cases are not very common, and I have found no instances in the records I have consulted. The best collection of such cases will be found in Mr. Paget's selies in the 'Lancet' of 1886. This collection contains five cases of parotitis following the introduction of instrumenlts into the bladder in patients with chronic cystitis, and one following the introduction of a pessary. The co-existence of rigors and high tempera- ture and the fatal result in three of the cases are in favour of the coinplication being pywmic. Two cases in MNIr. Paget's series followed operation on the cervix uteri. Dr. Goodell's paper read at a meeting of the American Gynme- cological Society and quoted by Mr. Paget, also contained cases of parotitis following operation for lacerated cervix, vesico-vaginal fistula, Tait's operation (two cases) and hysterectomy. When we find it recorded that both cases of Tait's operation and the one for vesico-vaginal fistula ended fatally, there is at least a suspicion of post-operative sepsis to account for the result. G(roup (3): the cases in connection with parturitioii anid abortion all appear to be essentially septic. AMy 66 SY31PTOMATIC PAROTITIS series contains two cases. Of the first case (No 71) I have few details. Unilateral parotitis developed during the puerperiumn and subsided without givinig much trouble. In the second case (No 72) the patient, in addition to paro- titis, developed parametritis and septic pneumonia. The temperature was of a hectic character and the case ended fatally. In Mr. Paget's series ('Lancet', 1886) will be found seven cases after delivery and abortion. Two cases developed peritonitis, two pyemia, and one had continued fever from uterine sepsis. The origin of the infection would seem to be as clear in these as in the post-operative cases. Condamin (26) has recently published an interestinig case of parotitis after abortion of a molar pregnancy. It was found necessary to curette the uterus for retained products of conception, and parotitis developed five davs after operation. The attack was preceded for four or five days by suppression of the salivary secretion. Mr. Paget found that puerperal parotitis usually developed some days later than post-operative. The reasoin for this, I should suggest, is that uterine infection frequently occurs some days after delivery, whilst in the post-operative cases peritoneal infectioni is usually either present prior to or occurs at the time of operation. Group (4): the parotitis that occurs in conniectioln with pregnancy and menstrual disorders is of a different order to the above and exhibits certain special character- istics. There are fewer of the attendant phenoimienia of inflammation with it, it never apparently goes on to suppura- tion, and it tends to be recurrent and periodic, recurring with pregnamncy or the menstrual epochs. The phenomena suggest a strong nervous elenment in its causation. 'rhe salivary glands are frequently affected during pregnancy, less commonly in connection with menstruation; but the affection is usually in the direction of increase or dimiinution in their secretory activity. Like the vomiting of pregnancy, it is to be regarded as a neurosis. In a case mentioned by Mr. Paget, in which abortion had to be SYMIPTOMATIC PAROTITIS 67 induced for continuous vomiting, salivation was a marked feature. Dr. Goodell has recorded two cases-one of hypersecretion of the salivary glands at each menstrual period, aind the other of a patient with a painful left ovary, in whom at every period the secretion of the left parotid was suppressed. The sympathetic contiection between the generative organs and the salivary glands is well illustrated by the above cases, but actual parotitis is rare under these conditions. A case of recurrent parotitis which first appeared in the second pregnancy and recurred regularly inl the five succeedinig ones is recorded *by Dr. Harkin (27). The left parotid was affected; it steadily enlarged during the pregnancy, and subsided after delivery, leaving a little permanent enlargement. One of my cases of parotitis (No. 68) occurred in connection with preg- nancy. The pregnancy was, however, complicated by chorea. The parotitis was certainly of the symptomatic, and not of the epidemic, variety. With regard to parotitis in connection with menstrual disorders Professor Peter's case is classical (28). It occurred in a young woman with amenorrhoea. At what should have been the periods there were firstly for two consecutive months attacks of parotitis, the next moonth swelling of the labia, then two more attacks of parotitis, and in the following month menstruation. Alr. Paget records that he found two instances of parotitis at the menopause. A very interesting case has been published recently by Dalche (29). The menopause in his patient was accompanied by much nervous disturbance. For a whole year the patient suffered from recurring attacks of parotitis, mostly at intervals of about a month, although the periods between them might be slightly longer or shorter. Both parotids were affected, with little or no fever, and without noticeable dryness of mouth. There is a case of Mr. Battle's that I shall quote later, where persistent VOL. LXXXVIII. 6 68 SYM1PTOMATIC PAROTITIS xerostomlia, with subsequent chronic enlargement of the parotids, followed the menopause. These cases are difficult to explain. The nervous system would appear to play an important r6ole in determining the attack, but it is difficult to see how nervous influences could do more than influence the secretion. There seems no reason to look on the under- lying cause as being septic, for the definite regularity of the attacks would be against an infective origin. Possibly, as Dalche suggests, the actual swellinig of the gland may be of vaso-motor origin, and we may adopt his view that the process is lighted up by a genital auto-intoxication, and provisionally classify this group as toxic.

III. PAROTITIS FOLLOWING OPERATION OR INJURY OF PORTIONS OF THE BoDY OTHER THAN THE ABDO31INAL AND PELVIC VISCERA. Examples of this are exceedingly rare, and the paro- titis is apparently always pyemic. Mr. Paget collected thirteen such cases, and in all of them the parotitis was part of a general pyamia. McDonald records a case after amputatioln of the leg. Cases No. 75 and 76 in mny series belong to this group. In case No. 76 the patient, a boy aged 13, had an exostosis of the humerus relnoved. In addition to a suppurative parotitis he developed multiple abscesses. The case ended fatally. For the notes of Case No. 75 I am indebted to the kindness of Mr. Edgecumbe Venning. The patient, an old gentleman aged 82, fell, striking the left side of the chest against a fender and fracturing two ribs. He recovered fronm the accident, but eleven days later developed left-sided parotitis, with elevation of tempera- ture. Although at first threatening to suppurate, the swelling subsided within the course of a few days. Eighteen days after the appearance of the parotitis a tender swelling, the size of a pigeon's egg, developed SYM1PTOMATIC PAROTITIS 69 over the right sterno-m-lastoid, also accomnpanied by febrile disturbance. This also subsided, but the patient gradually sank and died.

IV. PAROTITIS IN CONNECTION WITH DISEASES OF NIETABOLISM AND CHRONIC INTOXICATION. There is a variety of parotitis that occurs in colnnection with certain chronic metabolic diseases, such as gout and diabetes. Parotitis may also occur in cases of chronic poisoning with lead, copper and potassium iodide. I have grouped these together as cases of toxic parotitis. The cases are undoubtedly rare and have been described mostly by French observers.

Parotitis In Gout. Four cases of this rare complication of acute articular gout have been collected by Dr. Debout D'Estrees (30) in a paper read before this Society. One patient was under the care of Sir Archibald Garrod; the other cases were observed in France. In three of these cases there was a very remarkable sequence of events. The attacks began with swelling of one parotid, on the subsidence of which there was an acute gouty attack in a joint on the opposite side of the body; then in succession the opposite parotid an-d the corresponding joint to that first affected became involved. In one of the cases this sequence of events recurred five times in succession at intervals of one to two years. The fourth case differed from the others in that an attack of gout in the right wrist was followed by a painful swelling of the right parotid. The sequence of events in these patients would apparently leave no doubt that the same cause is at work in the case of the parotitis and the articular attacks, namely, irritation by urates. One of the patients experi- enced a saline taste in the mouth before meals, wNhich the 70 SYMPTOMATIC PARO'FITIS author considers to be due to the exeretioni of urates iil the saliva. I lhave recently beeni fortuniate enough to observe onie of these rare cases, in private practice. The patient (No 77 of series), anl alcoholic manl of 32, had had six or seven attacks of gout. The present attack commenced with involvement of the left anikle, the opposite ankle being attacked on the following day, and subsequently the left knee. Three days after the onset of the gout the left parotid became swollen anid painful, and two days later the right parotid became involved. The parotitis wvas associated with difficulty in swallowing and dryiness of mouth. The skin over the parotids was inflaimied anid edematous, resembling someiwhat the skin over a gouty joint. The duct papillw were niot inflamed and there was no saline taste in the mouth. The submaxillary glands were slightly enllarged anid tender, and there was some swellinig of the tissues beneath the chiin. The parotitis was not accompanied by any elevation of the temperature. The swelling of the glands lasted about nin-e days anid then subsided. As it developed the articular attack sub- sided, and on the resolution of the parotids there was a fresh outbreak in the feet. The parotitis was associated with peculiar attacks of paroxysmal pain, which, starting in the neck, extended upwards to the vertex and forwards to the parotid region and exterinal auditory meatus along the course of the second cervical nerve. 'l'hese attacks would recur once or twice an hour anid were particularly distressing to the patient. One of the cases in miiy series (No. 66) is interestiing in this connectioln, in that a septic suppurative arthritis in a gouty subject aged 75 was followed by suppurative parotitis of the opposite side as the sole secondary pyfemic complication. It suggests that there was some lowering of resistanice on the part of the parotid, whicl determined its selection by the micro-organisms. Another of my cases (No. 56) illustrates the occurrence of parotitis in a gouty patient aged 40, during the SYMPTOM31ATIC PAROTITIS 77]I terimiinal urwl'iia of chronic nephritis. There were nio other infective complications. Flexner has showni that terminial infections are commoon in granular kidney and ilmorbus cordis, anid the parotitis in this case appears to be of that nature. The selection of the parotid for the teriminal infectioln is interesting. I have also found parotitis occurring in patients with gonorrhoeal arthritis, but in this case it is probably part of a general pyemiia. Parotitis occurring in cases of chroniic lead-poisoning (saturnine parotitis) is probably closely allied to gouty parotitis. It usually occurs in conijunction with attacks of colic, and possibly the intestinal disturbance may predispose the gland to attack. A paper oni the subject has been published by Petit aind Hudelo (31) in con- nection with two cases that they exhibited before the Societe 31edicale des H(4pitaux, in Paris, in 1899. The attacks imay be recturrent, and may leave some permiianenlt enilargemiieint of the glan-d. In one of the above cases lead was detected in the saliva. The authors suggest that the parotitis may be due either to irritation of the gland by the lead during its excretion, or to a super- added infectioii of buccal origin. The first 'view would seeii the more probable, especially in view of the fact that the parotid affection tends in these cases to be recurrent or chronic. Chronic enlargement of the glands in plumbismn was described by Comby in 1882. A study of the subject will be found in 'La Presse Medicale' for December, 1897. Parotitis has also been described by Chauffard as occurrinig in connection with poisoning by copper, anid by several observers as followiing the admninistration of potassiuimi iodide.

V. PAROTIirs 1N THE SPECIFIC FEVERS. V&arious authors have drawn attentioni to the occurreince of parotitis as a complication of the following specific 7*2 SY3MPTOMIATIC PAPROTITIS fevers : typhus, enteric, pneumnonia, cholera, yellow and relapsinig fevers, seconidary syphilis, pynemia, and, rarely, smnallpox. The primuary diseases in which the cases in- my series occurred as a complication were enteric, pneu- monia, influenza, scarlet fever, facial erysipelas, gonorrhoea, and measles. IR Eqnter-ic Ferer. Parotitis is by no imieans a rare complication of this disease, its frequency varying somiiewhat in different epidemics. Mr. Paget, quoting the reports of the Lonidon Fever Hospital for the years 1870 to 1885, states that it occurred in 13 out of 2000 enteric patients, but only once am-longst 7000 cases of scarlet fever. According to Osler, there were 45 instances among the 2000 Muinich cases of enteric. Keen collected 75 cases of this coimpli- cation in enteric, and 353 in typhus, but the total number of cases amnong which these occurred is not stated. Osler states that it is less commonl in enteric than in typhus. Two cases in my series, both fatal, occurred in patients during the course of enteric fever. In the first case (No. 58), in addition to the parotitis there were other ter- minal infections-pericarditis and pleurisy. Post mortem, the Bacillus coli commugnis was isolated from the parotid, while pure cultures of the Bacilluts typhosus were grown from the spleen, liver, and bile. Although the presence of the Bacilluts coli communis does not exclude infection from the mouth, still, the well-known association of this organism with secon-dary typhoidal infection s is suggestive. In the seconid case (No. 59) the parotitis was also a terminal infection. There were no other mietastatic comiplications. The comparative frequency of parotitis inl enteric fever comnpared with that in the other commiion infectious diseases is interesting as shoWing the part played by lesions of the alimentary canal in disposing the gland to infection. In cholera, moreover, another disease with its local imiani- SY31PTOMATIC PAROTITIS 73 festationls in the intestine, parotitis is, accordinig to Osler, by no miieans uncommon. The fact, too, that parotitis is so much coimimoner in enteric than in scarlet fever, a disease associated with viruleint oral sepsis, would again poinlt to a metastatic origin of the infection rather than to a spread of infection by the duct.

Int Pneumonia. Three of the cases in iny series occurred in patients with lobar pneumonia, in one during the height of the disease, in the other two during convalescence. In the case in which it occurred during the height of the disease (No. 61) it was accompanied by pericarditis and inultiple abscesses, being distinctly pya-emic. The case was fatal, and the parotitis was again a termiiinal infection. In case No. 60 it developed on the eighth day of con- valescence, and although its onset was accompanied by a rapid rise of temperature from 98° to 104°, it resolved without suppuration- and did not hinder the patient's pro- gress. In case No. 62 it also developed on the eighth day of convalescence, and was accompanied by tonsillitis; there was no febrile disturbance with the attack, and it resolved without suppuration.

IbInflfuenza. Case No. 65 occurred in a patient with influenza, just at the height of the primary disease. It was unilateral and resolved rapidly.

In Scarlet Fever. Although this disease is one rich in metastatic com- plications, parotitis is very uncommon, and there would seem to be no reason for regarding it, when it does occur, as other than metastatic. If infection occurred by the 74 SYMPTOMATIC PAROTITIS duct, it should be imore conmnmoii. One case in imiy series (No. 64) occurred in a patient with this disease. It developed about the sixth day of the disease, was unilateral, and resolved, without suppuration. The patient recovered. One case (No. 63) followed measles in a debilitated infant, but was simply part of a general pyemia, and was, in fact, the first indication of it. The parotitis proceeded to suppuration, and was followed by multiple abscesses and septic broncho-pneumonia, the case terminating fatally. The cases of parotitis occurring in facial erysipelas and goniorrhceal pyemia I shall refer to in the succeediing section. Parotitis, then, as a coinplication of the specific fevers, may occur either during the height of the fever or during convalescence. As a rule it is of very grave significanlce if occurring at the height of the disease, an-d is almost always pyeamic, pointing to a generalisationi of the infec- tion. It usually appears earlier than the other metastatic infections, and may be valuable from the point of view of prognosis. It is practically always suppurative under these conditions. During convalescence it is not a grave complication, and does not affect the course of the case. Cases occurring during convalescence should be distin- guished from superinfection with ordinary mumps, and the iost valuable diagnostic sign is the suppression of salivary secretion that accompanies and usually precedes the symptomatic form. The absence of history of possible infection, and the fact of the parotitis remaining confined to the patient, will help to exclude the epidemic variety.

VI. PAROTITIS IN CONNECTION WITH LOCAL INFLAMMATORY CONDITIONS OF THE FACE AND NECK. Four of the cases of parotitis in my series occurred in connection with local inflammatory conditions, and I have ventured to draw attention to them because I cannot find that the process has been noticed to any extent. The parotid involvement would seem to occur by direct SYM1PTO3MATIC PAROTIT7IS 75) spread of infectioni, anid the cases have consequently been grouped by themselves. Case No. 68 occurred in connection with erysipelas of the face in a patient suffering from gonorrhoea. He had had eight or nine gonorrhceal attacks aind the last two had been complicated by arthritis. The subsidence of the second attack of arthritis was followed by the appearanice of an erysipelatous patch which gradually spread over the face. Unilateral parotitis developed a week after the subsidence of the erysipelas, and lasted for about a fort- night. A similar case is recorded by the late Dr. S. 0. Habershoni (32) and quoted by Mr. Paget. The patienit exhibited symptoms of general septicwinia and the parotid suppurated. In both these cases the parotitis followed anl erysipelatous attack in the neighbourhood, and may be due to a direct spread of infection. On the other hand, it might be a py,emic manifestationi. Parotitis occurred as part of a genieral pyeeiic infection in one of the cases collected by Thayer and Lazear (33) and published in their paper on " Gonorrhoeal Endocarditis and Septicoemia." Case No. 67 of my series is an exainple of the occurrence of parotitis in connection with facial erysipelas. The patient was an elderly alcoholic man, and the erysipelas followed a wound above the orbit. There was enormous swelling and inflammation of all the salivary glands. Case No. 68 is a case of parotitis in connection with Ludwig's angina. In addition to involvemeent of the fauces and larynx, the patient presented on admissioni much inflammatory oedema of the superficial tissues beneath the chin and behind the rainus of the jaw on the left side. The left parotid was much enlarged, and the right was involved to a slighter extenit. Although the primary conditioni yielded rapidly to treatimlenit, the parotid enlargement was still present a month later. Case No 69 is a very interestinlg case, in which parotitis developed in connectioni with a peculiar skin eruption of the face. I am iindebted for the details to Dr. Des Voeux. The patient, a wom-lan of 29, had suffered for seveni years 76 7SYMPTOMATIC PAROTITIS froimi a peculiar skinl eruption, recurriing anniiually inI autulin or winiter and lasting for about six imionths. She had occasional attacks of joint pains. In the winter of 1901, with the rash, a painful swelling of the parotid developed on the right side, accompanied by soimie febrile disturbance. In the following winter a corresponding swelling of the left parotid appeared with the appearance of the rash on the left side of the face. The attacks left the parotids permanently enlarged. The case was ex- hibited beforethe Dermatological Society last February (34). The diagnosis of the skin eruption rested between Erv- theina mnultiforme and . In these cases it seems most reasonable to suppose that the parotitis is due to a direct spread of infectioni to the gland from neighbouring structures. It must be rememi- bered that the salivary glands have lymphatic glands embedded in them, which drain the superficial parts of the face and neck. In the case of erysipelas or cellu- litis of the facial structures, these glands may become infected, and it is conceivable that the infection might spread froimi them to the parenchyma of the salivary glands.

B. CHRONIC AND RECURRENT PAROTITIS. Most of the recurrent and chronic cases of parotitis belong to the toxic group. The best example of recur- rence is exhibited by the group that occurs in connection with imienstrual disturbances. In gouty aind saturnine parotitis there is a similar tendency to recurrence, and in the latter case the glands may become chronically enlarged, with occasionally acute exacerbations following exacerbations of the primary disease. Mr. Battle (35) has recorded an interesting case, in which persistent followed the menopause. After three years the parotids slowly enlarged and remained permaniently enlarged, with the occasional supervention of acute attacks. In one of my cases, where parotitis occurred in connec- SYMPTOMATIC PAROTITIS 77 tion with a curious eruption of the face, the glandular en- largeimient becaine chronic. Occasionally chronic enlarge- mnent occurs without assignable cause. Case 75 in miy series illILstrates this. It was a case of slow but progres- sive enlargemient of the glands in a man of 24, without obstruction of the duct, and with Ino suspicion of new growth. The enlargeinent was associated with xerostomia, but saliva wa-; secreted on the introduction of food into the mouth. A similar case of chronic enlargement associated with xerostoimia is recorded by Dr. ilerringham (36). In this case the enlargement followed an attack of coryza. Another case of chronic enlargement in a, man of 18 with muscular atrophy was exhibited by Dr. Mitchell Clarke at the Neurological Society in February, 1903. Two cases of recurrent parotitis with xerostoinia will be found in Hutchinson's ' Archives of Surgery' (40). In the first case the parotid secretion was very easily arrested and seemed to be influenced by various kinds of food; the patient had suffered for twenty years from attacks of swelling of the gland, recurring every few months. The attacks were associated with pain in the fingers, and there was a gouty family history. In the second case the attacks coincided with attacks of polyuria. Both cases occurred in female patients, and Hutchinson reinarks that this disease seems to be almost confined to the female sex. The pheniomena of the attacks suggest some interference with the vaso-motor mechanism of the gland, and there is a possibility that the exciting cause in these cases might be purely nervous in origin. SOMmE GENERAL POINTS IN CONNECTION WITH SYMPTOMATIC PAROTITIS. Syimiptoimatic parotitis is a coinplication which imay arise under such very varying conditions, and is indeed a disorder of such coilmplex pathogeny, that it is impossible to regard it as a specific disease. Consequently it would Imardy be profitable to discuss such considerations as an 78 SYMIPTOMA'lTIC PAROTITIS incubation period, for evsen in the septic cases we cani irarely be certain at what period the infection of the glands occurs. In the imajority of post-operative cases it occurs as a rule within a few days of operation: the greatest number in nmy series occurred on the third day, the same day as the greatest number in Mlorley's cases. I have niot founid it to appear earlier than 48 hours after operation. lf the onset is delayed for a considerable timiie after opera- tion some other determining cause, such as septic peri- tonitis, is to be looked for. In the cases of gastric ulcer, treated medically, its oniset did not seem to be connected with any recognisable factor. When other complications occur in additioni to parotitis, parotitis as a rule precedes them, and should suggest a watch being kept for such complications. It should also suggest the possibility of sepsis within the operation area. The durationi of the attack is very variable and depends on its severity. The swelling may be over and gone withiln 36 hours, or may persist for a month or more. It tends to persist longer, and sometimes to become chronic in the toxic cases. The severity of the attack is variable. The attack iimay be mild, with little febrile disturbance, and subside quickly, or it may be more severe and go on to suppuration. Occasionally the inflamimatory process may extend beyond the gland itself, involvinig the surrounding structures in a conidition of spreading cellulitis, even downi as far as the clavicle. Very rarely the inflammation may be so intense that the gland actually sloughs. This occurred in Mr. Dyball's case (loc. cit.) alnd he has collected two or three other instances of it. It is confined to debilitated indi- viduals, or to conditions of the severest type of infection, usually associated with septicemia. Febrile disturbance occurs in the mnajority of cases. It occurred in 21 out of 27 post-operative cases, in which the temperature was recorded. The attack may be attended by very high temperature and yet resolve, and the tempera- tuire is of little assistanice in determiining the questioni of SYMPTOMATIC PAROTITIS p-9 possible suppuration. In one case the temiiperature ran up 60 with the attack, but there was nio suppuration. The question of suppuration is an important but diffi- cult one. In my series the proportion of cases in which parotid abscess occurred was very low-15 only out of 76 acute cases, or about 20 per cent. Akmong the cases following abdominal operations 7 out of 34 suppurated, among cases of gastric ulcer treated medically only 2 out of 15. If the complication occurs during the height of a specific fever, it almost invariably suppurates. The tendency to suppura- tioIn varies with the primary condition: the gland is more likely to suppurate in a patient with suppurative peritonitis or pyawmia than in a patient with simple gastric ulcer. InIMr. Paget's series 57 percent. of the cases suppurated, but this series was a much earlier one than mine, and the cases exhibited a much higher proportion of sepsis. The pro- portion of suppuration in Morley's cases,40 percent., is inter- mediate between Mr. Paget's and mine. The fact that in large series of cases the percentage of cases suppurating varies with the amount of primary sepsis among the cases is in favour of the view that infection occurs by the blood- stream and not from the mouth. The occurrence of suppuratioin is of slightly graver prog- nostic import, but as Mr. Paget has remarked, the cases do not die because the parotid suppurates. The occur- rence of parotitis affects the prognosis only in so far as it is an induration of some primiary septic focus, or as a forerunner of other septic complications. It is worth recording that in three or four of the cases suppuration involved the lymphatic glands in the neigh- bourhood only, and not the main parotid. Side affected. -The actual figures for the side affected in the cases in my series in which it is recorded are: Right side only, 21; left side only, 22 ; bilateral, 26. When bilateral, the involvement is rarely synchronous, one side usually being affected a day or two earlier than the other. With a unilateral primary lesion there is a slightly greater tendency for the gland of the same side SYMPTOMATIC PAR01TITIS to be affected. Of gynfecological cases, in Nos 29, :31, and 33 of imiy series and in several published cases the parotitis was oln the samne side as the diseased uterinie appendages. In the published cases of parotitis compli- cating appendix abscess the parotitis most coimnlonly was either right-sided or commenced on the right side. There is also apparently a greater tendency for the left parotid to be affected after lesions of the anterior surface of the stomach, aind the right after lesions of the posterior surface. Itmust be understood that this merely expresses the general tendency, and is far from being constant. Sex and age.-Sex has apparenitly no influence in dis- posing to this complication. The same remark applies to age, except that the complication is not commoni in children, even after abdominal operations. Organismns present.-The literature of the subject contains comparatively little on the bacteriology of syimiptomatic parotitis. In Bunts' and Morley's cases the organism present was Staphylococcus pyoyeines aureus. The same orgainisml was present in three cases in my series-one following ovariotomy, one after suture of a gastric ulcer, and the third in a case of gastric ulcer treated medically. As already mentioned, in a case dying of enteric the Bacillus coli commutis was isolated from the gland. In all the cases the organisms were present in pure culture. In the cases of parotitis complicating appendix abscess published by Bunts the Staphylococcuts pyogenes was isolated in pure culture in one case, while in another there was a mixed infection with staphylococcus and the B. coli CoimtinUnis. Mlorbid antatorny.-The gland was examined miicroscopi- cally in the case of a patient dying of enteric fever (No ,58) by Dr. Salusbury Trevor. There was a patchy smiiall- celled infiltration betweeni the acini, with much fatty change in the cells of the parenchbyma. The changes corresponded closely to those in the ca,se described by Mr. Paget. The inflammatory changes seem to be chiefly interstitial in character. SYMPTOMATIC PAROTITIS 81 The differential d(agaosis of sy8iptonmatic ancd epidemic parotitis.-The question of the parotitis beinig unilateral or bilateral is of no value in the diagnosis. The most important point is the condition of the secretioni. It is usually little affected in ordinary mumps (Fagge and Pye- Smith), while in the symptomatic variety the gland swell- ing is constantly accompanied and usually preceded by suppression of secretion. The outlying structures aire more commiionly involved in the accompanying inflamma- tion in all but the mildest cases of symptomatic parotitis. Epidemic mumps is common in children, the syinptomilatic variety is uncommon. Genital metastasis does not occur in the symptomatic form, and the occurrence of suppura- tion is distinctive. The symptomatic variety is not in- fectious from one patient to another. Dr. Addenbrooke (37) has recorded a case in which parotitis following an operation for peritonlitis in a boy of 13 was fol- lowed by an outbreak of mumps in other members of the family. Such cases are open to suspicion of being instances of the epidemic variety, because, after all, there is no reason why epidemic parotitis should not occur in a patient after operation as well as at any other timne. Prophylactic treatment.-Condamin suggests the em- ployment of injections of pilocarpine, wxhen operatioin is followed by marked suppression of secretion. The use of antiseptic mouth-washes, with a view to combating oral sepsis, has in our experience had no effect whatever in the prevention of parotitis. At the same time, they do no harm. When parotitis has appeared, the ordilnary lead and opium lotion is useful as a local application. To emiiploy belladonna is most irrational, because it tends to still further paralyse secretion and increase the dryness of mouth which is onie of the most distressing features of this complication. The question of incising the swelling needs careful consideration. The large proportion of cases ending in suppuiration in the earlier series has led to a statement as to the niecessity for early incisioni of the 82 SYMIPTO3ATIC PAROTITIS gland. The results of my series do Inot support this line of treatment, and I would suggest that the surgeon wait for evidenice of suppuration. More particularly is ex- pectant treatment indicated wheni there are no signs of -sepsis elsewhere. It has been remarked, and not witlhout justice, that the surest way to encourage suppuration is to incise the gland. At the same time, in post-operative cases the swelling should be watched carefully, as it is not possible to tell which cases will end in abscess formation. If an abscess form and ani exit be not giveen for the pus, it sometimes burrows through into the external ear. Even when this has occurred, however, it has apparently been followed by no permanent ill effects. It should be remem'ibered that over and over again the most threatening cases resolve completely. CONCLUSION S. 1. That cases of acute symptomatic parotitis fall into two main groups-septic anid toxic. The septic group includes post-operative cases, cases occurrilng in connlection with lesions of the alimentary canal and its appendages, the infective fevers and general pyaimia, the puerperium, and cases arising by direct spread of infection in inflam- matory conditions of the head and neck. The toxic group consists of cases occurring in connection with diseases of metabolism, such as gout and certain chronic intoxications, as well as, in all probability, cases in connection with disturbances of menstruation. There is aniother group of cases, mostly recurrent, which appear to be largely deter- mined by vaso-miiotor disturbances. 2. That post-operative parotitis is almost enitirely confined, as a complication, to abdominal operatioins, being coimimoniest after operationis on the stomach and condi- tionis of profound gastro-intestinal disturbance, such as acute intestinal obstruction. 3. That it is a well-marked complication of gastric ulcer. 4. That it is not particularly common after operations on the pelvic organs. SYMPTOMATIC PAROTITIS 83 5. That damage to the alimentary canal disposes the gland to infection by suppressing its secretion; and that the disposing cause after operations on the pelvic organs is probably the accompanying damage to the alimentary canal and peritoneum. 6. That with a unilateral primary lesion the gland of the same side is on the whole more disposed to infe'ction. 7. That parotitis as a complication of abdominal lesions is almost always associated with some degree of intra- abdominal sepsis, more particularly peritonitis. 8. That it may be a solitary complication, but is fre- quently accompanied by other septic complications, and its severity varies with the severity of the latter, when presenit. 9. That the majority of cases tend to subside without abscess formation, and the tendency to suppuration depends on the degree of general sepsis. 10. That it does not depend on the absence of food from the mouth, and that the emiployment of oral anti- septics is of no value in preventing this complication. 11. That the path of infection is in the majority of cases by the blood-stream and not by the duct from the mouth. 12. That suppuration sometimes affects the lymphatic glanids in the neighbourhood of the parotid, without involving the gland itself. 13. That the organism usually present in the post- operative cases is the Staphylococcus pyogenes aureus. 14. That except in so far as it is an indication of sepsis it does not affect the prognosis of the case. 15. That it may occur in connection with the specific fevers, either during the height of the disease or during convalescence. If occurring during the height of the disease, it is a sign of a generalised inlfection, and is of grave import; during convalescence it is usually of slight importance. 16. That in cases occurring during pregnancy, or in connection with menstrual irregularities there is a strong nervous element, but the exciting cause is -possibly toxic. VOL. LXXXVIII. 7 84 SYMPTOMATIC PAROTITIS 17. That the toxic cases tend to recurrence and to chronic enlargement of the gland. 18. That sex and age have no influence on its inci- dence, except that it is uncommon in young children. 19. That the chief sign that distinguishes it from epidemic parotitis is a preceding anld accompanying sup- pression of secretioni. 20. That early incision of the gland in cases of parotitis is not indicated as a routine treatment.

REFERENCES. 1. PAGET.-Lancet, vol. i, 1886 (resume' of sixty cases), p. 731; Brit. Med. Journ., vol. i, 1887, pp. 332, 613; Med. Soc. Proc., Feb., 1887. 2. MORLEY.-'American Gynmecology, Dec., 1902. 3. ENGLISH.-Med.-Chir. Trans., vol. lxxxvii, 1903. 4. PHILLIPS AND SILCOCK.-Lancet, 1899, vol. i, p. 832. 5. AITKEN.-Brit. Med. Journ., March 19th, 1904, p. 665. 6. BLUMER.-Brit. Med. Journ., 1903, vol. i, p. 1087. 7. MANSELL MOULLIN.-Lancet, 1902, vol. ii, p. 17. 8. KNOWSLEY THORNTON.-Lancet, 1886, vol. i, p. 57. 9. PH1LLIPS.-Case of Thrombosis of Inferior Vena Cava, Clin. Soc. Trans., 1901, p. 24. 10. JEFFERIES.-Lanlcet, Dec. 12, 1903, p. 1648. 11. STANLEY THOMAS.-Lancet, 190l, vol. ii, p. 1796. 12. ELDER.-Lancet, 1901, vol. i, p. 176. 13. BARJON.-Mem. et Comptes Rend. de la Soc. des Sci. Med. de Lyon, 1898. 14. FISKE JONES.-Boston Med. and Surg. Journ., Nov., 1902. 15. DYBALL.-Brit. Med. Journ., April 30th, 1904. 16. KNOWSLEY THORNTON.-" Three Hundred Additional Cases of Ovariotomy," Med.-Chir. Trans., 1887, p. 41. 17. LEWERS.-Diseases of Women, 5th edition, p. 377. 18. MAYO ROBSON.-"Pathology and Surgery of certain Diseases of the Pancreas," Lancet, March 26th, 1904. SYMIPTOMATIC PAROTITIS 85 19. CUCHE.-Bull. de la Soc. Med. des Hlop., Mlarch 5th, 1897. 20. PRIESTLEY.-JOuIrn. of Amler. Med. Assoc., July 7th, 1900. 21. JACOB.-Med. Rev., vol. iii. 22. SIMONIN.-Bull de la Soc., Med. des HOp., July 30th, 1903, p. 928. 23. PEACOCKE.-Lancet, vol. i, 1902, p. 28. 24. ENRIQUEZ AND HALLION.-La Presse Medicale, Jan. 24th, 1903. 25. BAYLISS AND STARLING.-Mechanisni of Pancreatic Secretion, Journ. of Phys., Sept., 1902. 26. CONDAM1IN.-La Semaine Gynecologique, June 2nd, 1903. 27. HARKINT.-Lancet, Feb., 1886, p. 374. 28. PETER.-Gaz. des Holp., 1868. 29. DALCHE.-La Gynecologie, Aug., 1903. 30. DEBOUT D'ESTREES.-" Gouty Parotitis," Med.-Chir. Trans., 1887, p. 217. 31. PETIT AND HUDELO.-" Saturnine Parotitis," Bull. de la Soc. Med. des H6p., Dec. 8th, 1899. 32. HABERSHON.-GuIy's Hosp. Rep., 1861. 33. THAYER AND LAZEAR.-JOUIrn. Exper. MAed., Jan., 1899, Case o. 34. British Journal of Dermatology, Feb., 1904. 35. BATTLE.-Clin. Soc. Tranis., 1895, p. 282. 36. HERRINGHAM.-Clin. Soc. Trans., 1886, p. 306. 37. ADDENBROOKE.-Lancet, Dec. 29, 1900, p. 1873. 38. BUNTS.-Amer. Journi. MIed. Sci., May, 1904. :39. McDONALD.-Edin. Med. Journ., 1885, vol. xxx, p. 1020. 40. HUTCHINSON.-Archives of Surgery, vol. iii, 1891. 86 SYMPTOMATIC PAROTITIS

Post-operative Cases.

Feeding at Case. Sex. Age. Disease. Operation. onset. Date of onset. F. 26- 1 F. 26 Perforating gastric Suture None 3rd day ulcer, anterior sur- face 2 26 P rectal 7th day 3 JE.4.4 Rectal 3rd day 4 F. 30 Perforating gastric Mouth for 4 6th day ulcer,posterior sur- days I face 5 F. 29 Perforating gastric Rectal 3rd day ulcer, anterior sur- face 6 F. 25 Non-perforating By mouth 8 9th day ulcer days before onset 7 F. 42 Non-perf orating Rectal 2nd day ulcer, lesser curva- ture 8 F. 38 Non-perforat ing Excision ulcer, posterior sur- face 9 F. 41 Old gastric ulcer Division of ad- By mouth 2 3rd day hesions days 10 F. 42 Carcinoma of ceso- Gastrostomy By stomach 4th day phagus 11 M. |, 10th day 12 F. 40) Caircinoma pylori Posterior gastro- Mouth jejunostomy 13 M. 55 Anterior gastro- Mouth and 13th day jejunostomy rectum 14 M. 34 Posterior gastro- Mouth 3 days 6th day jejunostomy 15 M. 36 Volvulus Redluction 16 F. 49 Obstruction by band Division of band Rectal 4th day 17 F. 44 Mouth 2 days 3rd day 18 F., 56 Malignant obstruc- Colotomy Mouth 13th day tion 19 F. I'- Mouth and 5th day rectum 20 M. 38 Strangulated hernia Herniotomy, Mouth 8 days 12th day omentum re- before onset moved 21 F. 34 Mouth 3 days 4th day 22 F. '70 Mouth 5 days 5th day

23 M. 49 Herniotomy Mouth SYMPTOMATIC PAROTITIS 87

Post-operative Cases.

Suppu- Duration. Side.Sd. Temperaturelraised. ration. Other complications. Result.

1 month + R., then 20 vYTes Pneumonia, pulmonary Death. L. abscess, empyema

3 days L. nil No Pleurisy Recovery. 6 days R., then To 102° Pneumonia L. 3 days ,,.y nil Pleurisy each side ,, ~~~~~?

20 days R. 30

Till death 30 IYTes Death 8th day ,, after operation.

4 days ,, 2 30 No Recovery.

? L. 30

3 days nil Till death R. ? Death. ? ?p P Thrombosis Recovery. 9-10 days L., then 20 R. 1 month + L. Temperature IYes irregular Till death L. 30 No General peritonitis Death 5 days after operation. I 1 month R. 30 Yes Suppuration of wound Recovery. A few days ,, nil No _ Till death raised Pleurisy Death.

p L. P Thrombosis Recovery.

7 days R. 10 Pleural effusion ,, Till death R., then To 102' Pneumonia Death 10 days L. after operation. 36 hours Symmet- nil Recovery. rical 88 SYMPTOMATIC PAROTITIS Post-operative Cases-continued.

Case. Sex. Age. Disease. Operation. onFeedng. at Dte ofot. .set.

24 F. 44 Ventral hernia Sac excised Mouth ? 25 F. 27 Tuberculous perito- Laparotomy ,, 2nd da',y nitis 26 F. 37 Cholelithiasis 6th da' 27 F. 16 Pelvic neurosis Appendicectomy Fluids hby 5th da mouth 28 F. 42 Ectopic gestation Laparotomy, Mouth 7 weeks aifter placenta left operatic yn 29 F. 29 Ovarian cyst Ovariotomy Mouth and 3rd da: rectumi 30 F. ,, ,, Mouth 2nd da y 31 F. - ,, ,, ,, 20th da1.7

32 F. 35 ,, ,, Mouth 2 days 3rd da;,y 33 F. 31 Hydrosalpinx Removal of Mouth and appendages rectum 34 F. - Malignant ovarian Exploratory lapa- Mouth 3 weeks a,fter cyst rotomy, tapping operati(on of cyst

Analysis of 34 post-operative cases. 23 recovered, the gland suppurating in 5 and resolving in eighteen cases. 11 died, ,, ,, 2. Non-ope.ative Abdominal Cases.

Case. Sex. Age. Disease. Feeding at onset. Date of onsget.|

35 F. 48 Gastric ulcer Mouth 10 days 22nd da3 36 F. 27 ,, Rectum 5th day 37 F. 23 ,, ,, 15th da3 38 F. 21 ,, ,, 14th da3 39 F. 16 ,, ,, 8th day 40 F. 22 Mouth Before admiIssion 41 F. 33 ,, Mouth 3 days 15th da3Y 42 F. 34 ,, Rectum 3rd day 43 F. 41 ,, ,, 9th day 44 F. 19 ,, Mouth and rectum 14th da3 45 F. 29 ,, Mouth 3 days 15th da3 46 F. 24 ,, Mouth 1 day 9th day SYMPTOMATIC PAROTITIS 89 Post-operative Cases-continued. u Temperature Duration. Side. raised. Srtiopu Other complications. Result.

2 ? Yes Abscess of palate, sepsis Recovery of wound 2 days Double 102-4e No ,, L. 10 ,, 7 days To 1020 1 month+ R., then ? Peritonitis (before paro- Death. L. titis) Till death R. 20 Intestinal obstruction Death 10 days after operation. Double 30 Pelvic peritonitis Death 6 days after operation. 1,, L. Irregular Peritonitis Death 24 days after operation. 3 weeks + L., then 20 Yes Pelvic sepsis, signs of Recovery. R. hepatic abscess 1 month + R. 10 No Till death nil Death.

Feeding by mouth, 23 cases; by mouth and rectum, 4 cases; by rectum, 6 cases; starved, 1 case.

Non-operative Abdominal Gases.

Duration. Side. Temperature Result. raised. ration. Other complications.

: ? ~~L.? No - Recovery. 19 days L., then To 1020 ,,~1 - R. 11 days R. I? Death. 6 days R. Recovery. Till death L., then 10 Y,, Death. R. 2 to 3 days L. ? Recovery. 11 days R. 20 ,,,UrticariaI ,1. 22cdays R. 20 Yes ,,

? L. 2°20 ,, 6 days 2? No ,,5 4 days L., then 30 P,, R. ,, ? L., then 20 Thrombosis R. P,, ,, 90 SYMPTOMATIC PAROTITIS Non-operative Abdominal Cases-continued.

Case. Sex. Age. Disease. Feeding at onset. Date of onset.

47 F. 24 Gastric ulcer Mouth 6 days 7th day

48 F. 26 Mouth 11 days 20th day 49 F. 24 Mouth 2 days 6th day 50 M. ? Duodenal ulcer Mouth 7 days 51 F. 43 Pelvic peritonitis 7 days 52 F. 29 Ulcerative colitis Rectum 3 days 53 F. 48 Ulceration of intestine Mouth 10 days 54 F. 58 Cholelithiasis 8 days 55 F. 51 6 days 56 M. 46 Ursemia and hemor- 6 days rhage 57 F. 4 mos. Gastro.enteritis 4 days

Specific Fevers, etc. 58 M. 28 Enteric Mouth 28th day of disease 59 M. 17 ,, 20th day of disease 60 M. Pneumonia 16 days after admission during, convalescence 61 M. 36 8th day of disease 62 M. 22 9 days after fall of temperature in convalescence 63 F. 6 mos. Measles After attack, pywemic 64 F. 21 Scarlet feveifr 5th day of disease 65 M. 20 Influenza At height of disease 66 M. 75 Pytemia 6th day after admission 67 M. Erysipelas 68 M. 34 Erysipelas, Eronorrhceal, 13 days after arthritis onset of ery-, sipelas

69 M. - Angina ludov. With angina SYMPTOMATIC PAROTITIS 91 Non-operative Abdominal Cases-continued.

Temperature IDuratiorn. Side. raised. ration. Other complications. Result. r I ns 2 weeks L. To 103-6 Yes,of - Recovery. lymphl glands 5 days R. None No - 9,, 1 week L. 10 ", Till death R. 30 Yes - Death. 1 month + R. 10 No Recovery. 12 days Double 30 ,, Death. 8 days R., then 20 n) R e L.

R. None ,, 5 weeks + L. To 1030 Yes Multiple abscesses Recovery. Till death L. 2e No Death. L., then Irregular Yes Bronchitis Recovery. R. Spec,ific Fevers, etc. Till death R. Slight rise No Pericarditis, pleural effu- Death next day. sion, acute nephritis ,, Double ,,_ Death 2 days ,, after onset. 5 days L. Nil Acute nephritis Recovery.

I~~~ ~ ~

Till death , p Yes Multiple abscesses, peri- Death 5 days carditis after. 4-5 days Double No - Recovery.

L. Yes Multiple abscesses Death. p No Recovery. 2-3 days ? iTill death L. Yes Pya?mic arthritis Death. ,, Double No 13 days R. Recovery.ee

1 month Double 92 9 SYMPTOMATIC PAROTJTIS. Various Diseases.

ICase. Sex. Age. Disease. Feeding at onset. Date of onset.

70 F. 29 Pregnancy and chorea Mouth 13 days after admission 71 F. 1? Puerperium 72 F. 32 3 days after delivery 73 F. 30 Erytheiiia multiforne With rash

74 M. 24 Chronic enlargement of parotids 73 M. 82 Fracture of ribs, left , 11 days after side injury 76 NI. 13 Following excision of exostosis 77 M. 52 Gout 3 days after articular attack SYMPTOMATIC PAROTITIS. 93 Various Diseases.

Duration. Side. Temperatureraised. Suppu-ration. OhrcmlctosOther complicatio Result.eut

14 days Double No Herpes Recovery. A fewdays ? , Subsided R. Septic Parametritis, pneumonia Death. after 4-5 days Recurrent Double Raised ,, Occasional articular at- Recovery. and tacks chronic t

A few days L Raised ,, Swelling over right Death. sterno-mastoid ?? Raised Yes Multiple pyaemic ab- scesses 9 days Double Unaffected No - Recovery. 94 SYMPTOMIATIC PAROTITIS

DISCUSSION The PRESIDENT, in thanking the authors of the papers for their initerestilng colmmiiiunications, remarked that they were at onle in holding the view of the secondary infection in symiiptomatic parotitis, yet held colntrary views as to the path of infection, the one holding that it was through Stenson's duct, and the other through the blood-stream, the conditions for infection being founid in the funletional inactivity of the duct. He hoped they should hear the experiences of those present as to how far one or the other of these views seemed to be valid, alnd whether in different cases olne or the other might not be applicable. Mr. STEPHEN PAGET alluded to the series of cases published l)y him miialny years ago, and referred to in the papers. Oni one poinit he desired to take exceptioni to statemiients in one or other of the papers; in regard to treatmenit he considered that a tiiiiely incisioln was indicated in the suspected presence of pus folrmiation, anid short of this he believed that leeching anid belladonina fomiienitationis were valuable. The cause of all these cases of parotitis was clearlv infection, and the view previously held as to the possibility of a sympathetic parotitis was now untenable. The sections which Mr. Bucknall had exhibited seemiied to him to prove almost conclusively that inifection took place through the duct, but at the samiie time he had to admit that Dr. Tebbs in his paper had brought forward considerable evidence pointing to an ilnfectioni by miiealns of the blood-stream, a reflex inhibition of the secretioni of the gland being a predisposing cause. If the parotitis were held to be onily due to inifectioni by the duct in a debilitated subject, the question arose, Why was it lnot more commii1ioni, seeing that all the necessary conditions for the occur- rence of duct inifection were so frequent? Mr. J. D. MALCOLM also thought that the microscopic sections slhowni by Mr. Bucknall indicated clearly that, in the cases from which these sections were obtained, infection of the glanld took place through the duct. He thought that the iml-portaince of septic mischief ill coinnection with an operationi, as a cause of this complication, had been exaggerated. His experience was all associated with abdominial operationis, alnd he had seen parotitis wheni there was lno other cause of trouble after a simlple ovari- otomy, sometimes as late as three weeks after the operation. He lhad poilnted out ('Brit. Med. Journ.,' 1899, vol. ii, p. 1673) that, after an abdomilnal section, the dorsal position, the liquid food, alnd consequent absence of pressure of the muscles of the gland anld duct durilng masticating, the febrile colndition, and in former times the free use of opium, all tended to dry the muouth and so SYMPTOMATIC PAROTITIS 95 favoured infection through the duct. In recenit vears he had seen this complication less frequently, but modern treatment, in addition to diminishing the risk of sepsis, allowecl the patients to take liquids freely and to take food which required chewing early, and it did not produce dryness of the mouth by the admiinis- tration of opiates. He referred to two cases of ovariotoiny in which the opening of Stenson's duct was found inflamed as soon as there was the slightest evidence of mischief in the . Mr. W. MCADAM ECCLES thought that both the views advanced as to the path of infection in cases of secondarv parotitis were very likely true, and he compared the infection of the parotid with that of the testicle in urethral inflammation, when the secondary epididymitis might be caused, he believed, either by direct extension or bv the blood-stream. He had seen cases of parotitis from both causes. He alluded to two cases of abdo- minal disease that had been under his observation, in whiclh, with no operation, parotitis had supervened after severe haenmor- rhage into the intestinal tract. In both cases the temperature had been raised before the onset of the attack, and suppuration occurred in the gland. Dr. C. 0. HAWTHORNE said that he considered the occurrence of parotitis in cases of gastric ulcer was very rare, and lie was surprised at the number of cases in which it had occurred in Dr. Tebbs' series. He mentioned two cases of parotitis follow- ing heematemesis which he had recorded, and he could only find reference to one other case, and that was one included i'n Mr. Stephen Paget's paper. He had been of opinion that infection came by the blood-stream, but the specimens which had beeni shown by Mr. Bucknall seemed to prove that the infection in some cases, at all events, took place along the duct. He could not, however, understand whv these cases were liot of greater frequencv if infection was dependent on the condition of the mouth. He suggested that the condition might be due to solmie toxin, and compared it to the swelling of the gland which was sometimes produced by the administration of potassiumii iodide. Mr. BUCKNALL, in reply, said that the cases advanced in favour of the view of infection by the blood-stream, as in septic peritonitis,, might equally well be quoted in support of duct infection, as in them the mouth became very septic. In typhoid the lmlouth was usuallv much more affected in the way of ulcera- tion than in diphtheria, hence the greater frequency of parotitis in that dlisease. In some cases where there miiight be a very septic state of the mouth, the disease might directly increase thle secretions from the gland, and thus tend to prevent duct infee- tion and parotitis. If infection were bv the capillaries, it would be expected that thev would be thrombosed and prominient in the sections, which was not so. 96 SYMPTOMATIC PAROTITIS Dr. TEBBS thought: (1) The fact that the same organism was present in the gland, the duct, and the mouth, might be explained by the elimination of the organisms by the gland; (2) the pre- ponderance of organisms in the inspissated secretion blocking the ducts would be accounted for by their being situated in a favourable culture miiedium, so that they would accumulate here rather than in the peripheral parts of the lobules, where they would be exposed to the bactericidal action of living gland cells; (3) that the argumeent deduced from the organism found in the gland being sometimes different to that of the primary disease might be subverted by the fact that when an infection became generalised the organisms of the secondary lesions were frequently different to those of the primary disease.